From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx2.suse.de ([195.135.220.15]:49088 "EHLO mx1.suse.de" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1750823AbdFTSiL (ORCPT ); Tue, 20 Jun 2017 14:38:11 -0400 Date: Tue, 20 Jun 2017 20:38:09 +0200 From: "Luis R. Rodriguez" Subject: Re: [PATCH 2/2] xfs: Properly retry failed inode items in case of error during buffer writeback Message-ID: <20170620183809.GV21846@wotan.suse.de> References: <20170616105445.3314-1-cmaiolino@redhat.com> <20170616105445.3314-3-cmaiolino@redhat.com> <20170616183510.GC21846@wotan.suse.de> <20170616192445.GG5421@birch.djwong.org> <20170616193755.GD21846@wotan.suse.de> <3ff0e0c8-ef9c-b7f8-d37e-ed02e5766c40@sandeen.net> <20170619105904.GA25255@bfoster.bfoster> <20170620165204.GP21846@wotan.suse.de> <20170620172041.GD3348@bfoster.bfoster> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170620172041.GD3348@bfoster.bfoster> Sender: linux-xfs-owner@vger.kernel.org List-ID: List-Id: xfs To: Brian Foster Cc: "Luis R. Rodriguez" , Eric Sandeen , "Darrick J. Wong" , Carlos Maiolino , linux-xfs@vger.kernel.org On Tue, Jun 20, 2017 at 01:20:41PM -0400, Brian Foster wrote: > On Tue, Jun 20, 2017 at 06:52:04PM +0200, Luis R. Rodriguez wrote: > > On Mon, Jun 19, 2017 at 06:59:05AM -0400, Brian Foster wrote: > > > If we truly needed a stable worthy fix in short order, that would > > > probably be to revert ac8809f9a ("xfs: abort metadata writeback on > > > permanent errors"), which caused this regression by making the AIL > > > responsible for failed retries. > > > > Should the following tag be added then to this commit proposed by Carlos: > > > > Fixes: ac8809f9a ("xfs: abort metadata writeback on permanent errors") > > > > That seems reasonable to me. Then in this case I'd like the commit log to also explain *why* the described fix did not work. It actually describes the issue as being considered, "thin provisioned devices that have run out of backing space", and this recovering. So did recovery never really work? Does recovery actually work for some cases but not some? If so why not for some? Luis